Yesterday's post was about Sahm's rule, which uses the unemployment rate to predict recessions. Friday's post was about the percentage of 25-54 year-olds employed.
It has been doing well while the unemployment rate has been going up a bit.The percentage of 25-54 year-olds employed is up 0.5 percentage points since December and has not declined for 7 straight months. Its average over the first 7 months of this year is slightly higher than both the 12 month average of last year and the average over the final 7 months of 2023.
Sahm's rule covers all the recessions since 1970. In all 8 of them, the percentage of 25-54 year-olds employed fell in at least one of the previous seven months (right now, as mentioned, that has not happened). In one case it fell 3 times and in another 4 times.
But, in some cases, the percentage of 25-54 year-olds employed rose over the entire 7 month period before the recession started. The list below has when each recession started followed by how much the percentage of 25-54 year-olds employed rose over the preceding 7 months.
December 1969 +0.3
November 1973 +0.7
January 1980 +0.7
July 1981 +0.5
July 1990 -0.4
March 2001 No change
December 2007 -0.3
February 2020 +0.9
In 4 cases (the first 4), it actually rose. So it seems that going up over a 7 month period is not a guarantee that a recession is not about to happen. But for two of the next 3 it did fall and for the other there was no change. The last one was the Covid recession which was highly unusual in that businesses simply closed all across the economy (at least, in part, due to government restrictions).
One last thing. I looked at all the cases since 1948 when the percentage of 25-54 year-olds employed had no declines for periods of 6, 7 and 8 months. I could have just done 7 months (since that is where we are at right now). But that is a pretty restrictive sample. This gives it a slightly broader range, but not too broad since this group is plus or minus one month from 7. So similar cases.
The first column tells us which cases in terms of months of no declines. The next table tells us when the last month in the streak was. Then the next column tells us when the next recession began after the streak ended. The last column tells us how long it took from the end of the streak until that next recession started.
|
Months of No Declines |
Date Period Ended |
Next Recession Began |
Lag in Months |
|
6 |
Aug-65 |
Dec-69 |
52 |
|
6 |
Oct-69 |
Dec-69 |
2 |
|
6 |
Jul-73 |
Nov-73 |
4 |
|
6 |
Aug-93 |
Mar-01 |
91 |
|
6 |
Dec-94 |
Mar-01 |
75 |
|
6 |
Oct-06 |
Dec-07 |
14 |
|
6 |
Apr-14 |
Feb-20 |
70 |
|
6 |
Mar-16 |
Feb-20 |
47 |
|
6 |
Jan-20 |
Feb-20 |
|
|
7 |
Jan-59 |
Apr-60 |
15 |
|
7 |
Jul-63 |
Dec-69 |
77 |
|
7 |
Jun-76 |
Jan-80 |
43 |
|
7 |
Mar-79 |
Jan-80 |
10 |
|
7 |
Apr-85 |
Jul-90 |
63 |
|
7 |
Jan-86 |
Jul-90 |
54 |
|
7 |
Jul-24 |
Unknown |
|
|
8 |
May-66 |
Dec-69 |
43 |
|
8 |
Oct-67 |
Dec-69 |
26 |
|
8 |
Jan-89 |
Jul-90 |
18 |
|
8 |
Jan-01 |
Mar-01 |
2 |
|
8 |
May-17 |
Feb-20 |
33 |
|
8 |
Aug-23 |
Unknown |
|
There are 22 cases, but two of them are so recent that then next recession has not yet happened. Of the other 20 cases, in 16 of them it was at least 12 months until the next recession began. The four lowest were 2, 2, 4 and 10 months later. The average lag until a recession began was about 36 months.
If we only look at the 6 cases when the percentage of 25-54 year-olds employed did not fall for exactly 7 straight months (the case we are in right now), the earliest a recession happened was 10 months later. Some were several years later.
See US Business Cycle Expansions and Contractions from The National Bureau of Economic Research.
Also see Employment-Population Ratio - 25-54 Yrs. from The Federal Reserve Bank of St. Louis.
No comments:
Post a Comment